Download Sensing earth`s rotation with a helium–neon ring laser operating at

April 15, 2015 / Vol. 40, No. 8 / OPTICS LETTERS
1705
Sensing earth’s rotation with a helium–neon ring laser
operating at 1.15 μm
K. Ulrich Schreiber,1,* Robert J. Thirkettle,2 Robert B. Hurst,2 David Follman,3 Garrett D. Cole,3,4
Markus Aspelmeyer,5 and Jon-Paul R. Wells6
1
Technische Universitaet Muenchen, Forschungseinrichtung Satellitengeodaesie, Geodaetisches Observatorium Wettzell,
93444 Bad Kötzting, Germany
2
Department of Physics and Astronomy, University of Canterbury, Private Bag 4800, Christchurch 8020, New Zealand
3
5
Crystalline Mirror Solutions LLC, 114 E Haley, Suite N, Santa Barbara, California 93101, USA
4
Crystalline Mirror Solutions GmbH, Seestadtstr. 27, Top 1.05, A-1220 Vienna, Austria
Vienna Center for Quantum Science and Technology (VCQ), Faculty of Physics, Univ. of Vienna, A-1090 Vienna, Austria
6
The Dodd-Walls Centre for Photonic and Quantum Technologies and Department of Physics and Astronomy,
University of Canterbury, Private Bag 4800, Christchurch 8020, New Zealand
*Corresponding author: [email protected]
Received February 13, 2015; revised March 8, 2015; accepted March 13, 2015;
posted March 16, 2015 (Doc. ID 234633); published April 8, 2015
We report on the operation of a 2.56 m2 helium–neon based ring laser interferometer at a wavelength of 1.152276 μm
using crystalline coated intracavity supermirrors. This work represents the first implementation of crystalline coatings in an active laser system and expands the core application area of these low-thermal-noise cavity end mirrors to
inertial sensing systems. Stable gyroscopic behavior can only be obtained with the addition of helium to the gain
medium as this quenches the 1.152502 μm (2s4 → 2p7 ) transition of the neon doublet which otherwise gives rise to
mode competition. For the first time at this wavelength, the ring laser is observed to readily unlock on the bias
provided by the earth’s rotation alone, yielding a Sagnac frequency of approximately 59 Hz. © 2015 Optical
Society of America
OCIS codes: (120.5790) Sagnac effect; (140.3370) Laser gyroscopes; (140.3560) Lasers, ring.
http://dx.doi.org/10.1364/OL.40.001705
The first demonstration of rotation sensing with a ring
laser gyroscope was performed by Macek and Davies
in 1963 using a 1 m2 helium–neon laser system operating
at a wavelength of 1.153 μm [1], just two years after the
first report of laser oscillation in a helium–neon discharge [2]. Their device utilized four gain tubes, one in
each of the sides of the square ring and, therefore, had
a total of 20 intracavity surfaces. Rotation sensing was
attained through an externally imposed rotation via a
mechanical turntable.
Over the last two and a half decades, large ring laser
gyroscopes have increased in size from approximately
1 m2 to over 800 m2 [3,4]. The reason for this is obvious
from Eq. (1) for laser gyroscopes, which relates the beat
frequency δf of the two counter propagating continuous
wave (CW) laser beams inside the ring cavity to the rate
of rotation (Ω) imposed upon the projection on the normal vector (n) of the ring laser structure [5,6]:
δf 4A
n · Ω:
λP
(1)
Thus, it is clear that increasing the perimeter (P) of the
device and, therefore, the area (A) enclosed by the two
laser beams leads to an increased sensor resolution.
Currently the highest usable sensitivity of 1 × 10−8 of the
earth’s rotation rate (ΩE ) is achieved by the German
Gross ring [7]. Ring lasers are now poised for realistic
attempts at terrestrial measurement of general relavistic
phenomena, such as the Lense–Thirring effect [8].
At some level of performance, all ring laser gyroscopes
are limited by geometrical instabilities, which manifest
themselves in the measurement quantity (δf ) through the
backscatter process. Backscatter arises from scattering
0146-9592/15/081705-04$15.00/0
of the intracavity laser beams into the counter propagating beam path, inducing both phase and optical frequency variations to that beam. Geometric instabilities
cause time-varying optical frequency fluctuations, which
are transferred to the Sagnac frequency as a time-varying
readout error superimposed on the signal from earth’s
rotation. In our gas lasers, the only intracavity elements
are the supermirrors which form the cavity itself and,
therefore, it is mirror imperfections that drive the backscatter process. One approach to minimize the influence
of backscatter is to utilize longer wavelength laser radiation since the backscatter amplitude will decrease dramatically, in the limit of Rayleigh scattering proportional
to the reciprocal of the fourth power of the laser wavelength [9]. The downside of such an approach is an
unavoidable reduction in the scale factor of the gyroscope. In principle, however, a net improvement in performance can be expected.
The supermirrors employed in this experiment consist
of 8 mm diameter crystalline coatings transferred to
super-polished fused silica substrates with a 4 m radius
of curvature (ROC). As in a previous demonstration of an
ultrastable optical reference cavity [10], these mirrors are
produced via a microfabrication-based substrate-transfer
technique, whereby a single-crystal Bragg mirror is initially grown on a lattice-matched GaAs wafer by molecular beam epitaxy (MBE), selectively removed from this
initial growth template using a series of lithography and
etching steps, and is then bonded to the final optic [Fig. 1
(inset)]. Such mirrors have the additional advantage of
minimizing thermally induced mechanical fluctuations
in the mirror coatings. This is unlike the industry standard ion beam sputtered dielectric multilayer mirrors
© 2015 Optical Society of America
1706
OPTICS LETTERS / Vol. 40, No. 8 / April 15, 2015
Fig. 1. Details of the crystalline supermirrors. The plot shows
the transmission spectrum for the ideal design and as-grown
multilayer under illumination at a 45° angle of incidence. The
inset includes a photograph of a completed mirror structure
consisting of an 8 mm diameter GaAs/AlGaAs mirror pad
transferred to a 25 mm diameter fused silica substrate. The
small flat on the coating disc indicates the crystal orientation,
in this case parallel to the h110i direction.
used in all previous ring lasers developed by our group.
Although small variations of cavity length should compensate for each other in an ideal situation, there are usually small noneciprocal effects that cause backscatter to
change and the beat note of the gyroscope to walk off
from its nominal frequency [11,12].
For single-mode operation in a ring configuration, the
coating is optimized for operation at a 45° angle of
incidence with a designed transmission of 1.7 and
13.6 ppm for s- and p-polarized light, respectively. In
detail, the high-reflectivity multilayer consists of 38.5
periods of alternating GaAs (high index, 3.45 at 1.152 μm,
85.5 nm ideal quarter-wave thickness) and Al0.92 Ga0.08 As
(low index, 2.96 at 1.152 μm, 99.5 nm ideal quarter-wave
thickness). Spectrophotometer measurements of the asgrown epitaxial films reveal a very small thickness error
of approximately −0.5%. The resulting 5.5 nm blue shift in
the center wavelength of the high-reflectivity stopband to
1.1445 μm yields transmission values of 1.9 ppm (s)
and 15.1 ppm (p) in the final optics at 1.152 μm.
Optical absorption in the coating was independently
probed on wafer at 1.155 μm via photothermal common
path interferometry [13]. Because of significant loss
within the optical path from water vapor in the ambient
laboratory environment (on the order of 15 ppm∕cm),
these measurements can only provide an upper limit
of 5 ppm, though there is evidence that the absorption
loss for these specific coatings is at the 1 ppm level.
At this early stage of development, scatter remains the
biggest limitation in terms of optical performance for the
crystalline supermirrors. The micro-roughness of the
multilayer was investigated via atomic force microscopy
and yielded an RMS value of 1.7 Å. In addition to the limiting micro-roughness, growth related defects dominate
the overall scatter losses in this experiment. These defects are known to occur in MBE-grown GaAs films and
work is currently under way to mitigate their presence.
For the employed coatings, we observe 10 μm sized scattering centers with a density on the order of 1000 cm−2 ,
while defect densities as low as 10 cm−2 have been
achieved in other instances. Note also that the large
optical beam diameter of 3.67 mm in this experiment
significantly increases the impact of scattering when
compared with a diameter of 0.5 mm with the previously
constructed linear reference cavity at 1.064 μm, where
total scatter losses below 5 ppm have been observed.
Using a simple exponential model [14] and ignoring interfacial or interference effects within the Bragg stack
yields a lower limit of 3.4 ppm for the optical scatter at
the operating wavelength of 1.152 μm, which is comparable to state-of-the-art ion beam sputtered supermirror
coatings.
Employing these novel coatings, we present the results
of experiments using a large helium–neon based ring laser operating in the near infrared. We have performed
our experiments on a vertically (wall) mounted, 6.4 m
perimeter square ring laser located in a second floor laboratory of a high rise building on the Ilam campus of the
University of Canterbury in Christchurch, New Zealand.
The cavity is entirely filled with helium–neon gas, and the
only intra-avity elements are the spherical supermirrors.
Radio frequency excitation of the gain medium at 80 MHz
occurs within a narrow pyrex plasma tube of 4 mm diameter, thereby avoiding Langmuir flow. Operation on a single longitudinal mode is achieved through gain starvation
with servo control of the rf power, maintaining a stable
laser output. As expected, the crystalline mirrors only
supported s-polarization in the laser cavity. A peltier
cooled InGaAs detector was used to measure the combined output beams of the laser, which was performed
using a short beam path in open air. The beam combiner
itself was aligned using the residual sensitivity of Sibased CCD cameras at the laser wavelength, as well
as standard hand-held IR viewers. The ring was first filled
with a 50∶50 mixture of Ne20 and Ne22 . Simultaneous
laser oscillation was then observed on the well-known
neon doublet split by 51 GHz (2s2 → 2p4 at 1.152276 μm
and 2s4 → 2p7 at 1.152502 μm) [15]. By varying the total
pressure of this mixture and the rf excitation power, the
regime of maximum gain was identified at a pressure of
0.2 hPa. Figure 2(a) shows the dependence of the laser
gain over the total gas pressure in the cavity.
A ringdown measurement of the cavity provided a
value of τ 20 μs, corresponding to a quality factor of
Fig. 2. Maximum obtainable laser power from pure neon as a
function of total gas pressure (a). Gyroscope operation was
possible in the regime between 1 and 10 hPa (b) when the laser
gain from the rf excitation was reduced.
April 15, 2015 / Vol. 40, No. 8 / OPTICS LETTERS
Q 2πντ 3.2 × 1010 with ν the oscillation frequency of
the laser cavity. Thus, the total loss (L) of the cavity with
a free spectral range (FSR) of 46.875 MHz amounts to
L 1∕τ × FSR 0.0011. Similar seemingly high loss
has been observed previously in many of our large ring
lasers [4] and speculatively attributed to “waviness” in
the mirror profiles. In this case, there is additional loss
because of beam clipping, where light is transmitted past
the edges of the reflective coatings. The total loss is a
subject of ongoing investigation. With the application
of a spatial filter blocking the spilled light, we estimate
that this extra light leakage is about four times larger
than the light loss from the transmission through the
mirrors. Of course, one must bear in mind that the center
of the TEM00 mode was attenuated by a factor of approximately 106 by the high-reflectivity coatings. Beam
clipping losses can be easily remedied in future experiments as crystalline coatings of approximately 20 mm
diameter have already been demonstrated in the context
of the development of low-noise optics for interferometric gravitational wave detectors [16]. In a second step,
4
He was slowly added and the achievable laser power
for a constant rf excitation of 8 W was measured. This
is shown in Fig. 2(b). With only a small fraction of
helium added, a significant drop in beam power was observed. This is attributable to the fact that helium efficiently quenches the 1.152502 μm transition through an
increase in the 2p7 population [15].
Compared with the operation of the same cavity on a
single longitudinal laser mode at a wavelength of
632.8 nm, the circulating laser power was considerably
smaller. It required approximately 0.6 W of rf power to
achieve 4.7 nW of output power, corresponding to
2.6 mW of intracavity laser power. In the visible, the same
system operates at 2.5 W of rf providing 12 nW of optical
power corresponding to 60 mW of intracavity beam
power. It was also found that the multimode threshold
is much closer to the monomode threshold in the infrared
as opposed to the visible. Furthermore, it appears that
increasing the gas pressure in the cavity does not enlarge
the separation between the multimode threshold and the
monomode threshold via homogeneous line broadening
for operation at 1.152276 μm, whereas for operation on
the 632.8 nm transition this has a considerable influence.
Additionally, we observed a significant imbalance in
beam power between the clockwise and counter-clockwise beams of about 15%, which is about a factor of 2
larger then routinely observed with ion beam sputtered
mirrors.
Despite the comparatively low cavity Q, the laser operated as a gyroscope, unlocking on the rate bias provided by the earth rotation alone. To the best of our
knowledge, this is the first time this has been reported
in the literature. While the operation on a pure neon
gas mix was helpful to align the laser properly by providing a brighter beam, it was of no use with respect to the
operation as a gyroscope because of the presence of a
highly variable additional beat note when simultaneous
laser oscillation on both neon lines was observed. We
reasonably attribute this phenomenon to mode competition. Since helium efficiently quenches the 1.152502 μm
transition, the 1.152276 μm transition is the only remaining optical frequency and the gyroscope turned out to be
1707
operable as a rotation sensing device over the range of
1–10 hPa of helium.
Modulations are observed on the clockwise and
counterclockwise beams at the Sagnac frequency given
by Eq. (1), with fractional amplitudes m1 , m2 which vary,
but are typically 5%–8%. These modulations allow estimates of backscatter fractional amplitude coupling r 1;2
using r 1;2 m1;2 πδf P∕c, with δf and P as in Eq. (1) [17].
In our case, amplitude couplings in the order 0.2–0.3 ppm
are indicated. The expected backscatter induced perturbations are of the order m1 m2 δf ∕2 [17], in our case
0.1–0.2 Hz. The total fractional scattered intensity may
be estimated as r 2s ≈ r 2 16d2 ∕λ2 [18] for beam diameter
d. With d 3.6 mm this gives ≈10 ppm. As this is the estimated scattering from four mirrors, it is in fairly good
accord with the earlier estimated 3.4 ppm for a single mirror. The rotation rate threshold for lock-in of a ring laser
gyro may be estimated as ΩL cλ2 r s ∕32πAd [18] for
laser area A (with other symbols already defined). In our
case, it is calculated as 1.4 × 10−6 rad∕s which is ≈3% of
the projected earth rotation rate. Thus, the observed
absence of lock-in is expected.
Figure 3 shows a timeseries of about 45 min of the measurement of the earth’s rotation as one example. Neither
the laboratory nor the stainless steel ring laser structure
was temperature stabilized; therefore, the data in the plot
is highpass filtered to remove sensor drift. Apart from the
earth’s rotation, the measured rotation signal also shows
a 2.36 Hz frequency modulation caused by the fundamental rocking mode of the entire eight floor Rutherford
building about its long axis in the form of a small oscillation about the mean of the beat note. The frequency
splitting of the two counter propagating laser beams according to Eq. (1) was measured to be 59 2 Hz. With
the latitude of the laboratory at 43.52° south, the vertical
orientation of the ring laser structure on the laboratory
wall and an orientation of the short side of the building of
about 32° east of north, the expected Sagnac frequency
according to Eq. (1) is around 60 Hz. In the presence of
backscatter coupling from a concomitant frequency pulling, the observed data is in good agreement with expectation. Figure 4 shows the corresponding spectrum of the
interferogram. The main beat note induced by earth’s
rotation shows the two sidebands induced by the rocking
Fig. 3. Sagnac beat note from the earth’s rotation, detected at
a wavelength of λ 1.152 μm. The observation was highpass
filtered to compensate for variations caused by backscatter
pulling.
1708
OPTICS LETTERS / Vol. 40, No. 8 / April 15, 2015
Fig. 4. Spectrum of the measured rotational signal shows the
earth’s rotation as the dominant signal, together with frequency
modulation sidebands induced by the rocking of the building.
motion of the building at 2.36 Hz separation on either side
of the main peak.
In summary, we have successfully obtained unlocked
rotation sensing of a large HeNe ring laser gyroscope in
the infrared regime rate biased by the earth’s rotation
alone. Moreover, we observe encouraging performance
for the first implementation of crystalline coatings in
an active laser cavity. Future work will focus on improvements in the overall system, as well as on the ultimate
optical and thermo-mechanical performance of crystalline supermirrors and their impact on large-area and
high-sensitivity RLGs.
The authors acknowledge support from the German
national science foundation DFG under contract
Schr645/6-1 and assistance from the Marsden Fund of the
Royal Society of New Zealand under contract 10-UOC041. G. D. Cole acknowledges support from EURAMET/
EMRP (QESOCAS) and thanks Dr. A. Alexandrovski of
SPTS for performing optical absorption measurements.
A portion of this work was performed in the UCSB
Nanofabrication Facility.
References
1. W. M. Macek and D. T. M. Davies, Jr., App. Phys. Lett. 2, 67
(1963).
2. A. Javan, W. R. Bennett, Jr., and D. R. Herriott, Phys. Rev.
Lett. 6, 106 (1961).
3. K. U. Schreiber and J.-P. R. Wells, Rev. Sci. Instrum. 84,
041101 (2013).
4. R. B. Hurst, G. E. Stedman, K. U. Schreiber, R. J. Thirkettle,
R. D. Graham, N. Rabeendran, and J.-P. R. Wells, J. Appl.
Phys. 105, 113115 (2009).
5. S. F. Jacobs and R. Zanoni, Am. J. Phys. 50, 659
(1982).
6. R. B. Hurst, J.-P. R. Wells, and G. E. Stedman, J. Opt. A 9,
838 (2007).
7. K. U. Schreiber, T. Klügel, A. Velikoseltsev, W. Schlüter,
G. E. Stedman, and J.-P. R. Wells, Pure Appl. Geophys.
166, 1485 (2009).
8. N. Beverini, M. Allegrini, A. Beghi, J. Belfi, B. Bouhadef, M.
Calamai, G. Carelli, D. Cuccato, A. Di Virgilio, E. Maccioni,
A. Ortolan, A. Porzio, R. Santagata, S. Solimeno, and A.
Tartaglia, Laser Phys. 24, 074005 (2014).
9. Y. B. Band, Light and Matter (Wiley, 2006), p. 187.
10. G. D. Cole, W. Zhang, J. Ye, and M. Aspelmeyer, Nat.
Photonics 7, 644 (2013).
11. R. W. Dunn and A. R. Hosman, J. Appl. Phys. 116, 173109
(2014).
12. R. Rodloff, IEEE J. Quantum Electron. 23, 438 (1987).
13. A. Alexandrovski, M. Fejer, A. Markosian, and R. Route,
Proc. SPIE 7193, 71930D (2009).
14. H. Davies, Proc. IEEE 101, 209 (1954).
15. W. R. Bennett, Jr. and J. W. Knutson, Jr., Proc. IEEE 52, 861
(1964).
16. J. Steinlechner, I. W. Martin, A. Bell, G. Cole, J. Hough, S.
Penn, S. Rowan, and S. Steinlechner, “Mapping the Optical
Absorption of a Substrate-Transferred Crystalline AlGaAs
Coating at 1.5 μm,” December 2014, http://arxiv.org/abs/
1412.3627.
17. R. B. Hurst, N. Rabeendran, K. U. Schreiber, and J.-P. R.
Wells, Appl. Opt. 53, 7610 (2014).
18. F. Aronowitz, Laser Applications, M. Ross, ed. (Academic,
1971), Vol. I, p. 133